The Design Of Strapdown AHRS Based On ARM And DSP

With the development of technology, more and more people are familiar with the navigation. As strapdown inertial navigation system (SINS) could achieve autonomous navigation without external information, so the SINS became the main form in the field of navigation. The AHRS solution is the key technologies in the SINS, we can get the data parameters of the navigation through the attitude matrix, so the AHRS solution is the important work in the SINS.Therefore, building a computer-based strapdown AHRS is the key to build an efficient, small, and high reliability strapdown navigation system.Navigation computer is a typical embedded system, ARM and DSP are the major constituent parts in the embedded field, and the ARM which has good control and rich peripheral interfaces plays an important role in the embedded computing; the DSP has a large-scale computing capacity of data and a particular advantage in digital signal processing.In the embedded operating system, thanks to its open-source, a good core features and a wealth of software resources, embedded Linux operating system has become an emerging force.DSP/BIOS is a real-time operating system for TI DSP, so it is helpful for the development of the TI's DSP.The design aimed at the design of system, major work focused on the hardware design and system software integration. So the design used the quaternion attitude update algorithm for the realization of the algorithm.Therefore, the dissertation will build a dual-CPU embedded AHRS based ARM+DSP. This design completely takes advantage of the characteristics of ARM and DSP, and they could achieve high-speed switching through Dual-port RAM. we builded the Embedded Linux operating system for ARM and DSP/BIOS real time operating system for DSP, reduced the difficulty to develop applications and drivers. Experiment result has proved that this system structure has excellent performance, small size, high reliability, so it is very suitable for the strapdown AHRS which needs real-time computing with huge amount of data.